The plan by U.S. chipmaker Nvidia to invest up to $100 billion (about CZK 2 trillion) in OpenAI, the company behind the ChatGPT models, has stalled. According to The Wall Street Journal, some Nvidia employees expressed doubts about the deal, and negotiations did not progress beyond the initial stages.
Terms of the Original Deal
The companies announced a memorandum of understanding last September at Nvidia's headquarters in Santa Clara, California. Nvidia pledged to build at least 10 gigawatts of computing capacity for OpenAI while investing up to $100 billion to help OpenAI pay for the infrastructure. The deal stipulated that OpenAI would lease the chips from Nvidia.
At the time of the announcement, OpenAI expected negotiations to be completed within several weeks. The reality was different, however—the talks stalled at an early stage, and the original megadeal is now on hold.
Criticism from Jensen Huang
In recent months, Nvidia CEO Jensen Huang has privately emphasized to industry partners that the original $100 billion deal was nonbinding and had not been finalized. Huang also privately criticized what he described as a lack of discipline in OpenAI's approach to business and expressed concerns about the competition OpenAI faces from companies such as Google and Anthropic.
During a visit to Taipei on Saturday, Huang told reporters that Nvidia would be "absolutely involved" in OpenAI's latest funding round. "We will invest a great deal of money, probably the largest investment we have ever made," he said. When asked whether it would exceed $100 billion, he replied: "No, no, nothing like that."
An OpenAI spokesperson said: "Our teams are actively working on the details of our partnership. Nvidia's technology has been foundational to our breakthroughs from the beginning, powers our systems today, and will remain key as we scale what comes next." An Nvidia spokesperson said in an emailed statement to Reuters: "We have been OpenAI's preferred partner for the past 10 years. We look forward to continuing to work together."
New Talks on a Smaller Investment
While the September deal remains stalled, Nvidia is continuing talks on a separate major investment in OpenAI as part of the AI developer's effort to raise $100 billion in new capital from multiple partners. Companies discussing investments worth tens of billions of dollars include Amazon and SoftBank. The total funding round could reach $100 billion (CZK 2 trillion).
According to other sources, Nvidia is now discussing a significantly smaller investment of around $30 billion (CZK 615 billion), less than one-third of the amount originally planned.
OpenAI has signed a number of deals with chipmakers and cloud companies that helped fuel the global rise in stock markets. However, investors have begun to worry about the startup's ability to finance these deals, leading to a partial sell-off in related stocks. OpenAI CEO Sam Altman said that OpenAI's commitments arising from these deals total $1.4 trillion (CZK 28.7 trillion), more than 100 times the revenue the company had planned to generate last year.
According to an estimate by Gartner, spending on artificial intelligence investments and related infrastructure could climb to $2.5 trillion (CZK 51.2 trillion) this year, representing a 44% increase compared with 2025.
Warnings Against Overblown Promises
In recent months, warnings have emerged about a possible investment bubble caused by the overvaluation of stocks linked to artificial intelligence technology. According to an analysis by Der Standard, many promises made by executives of AI-focused startups are overblown, and reality is far removed from their plans. International Monetary Fund Managing Director Kristalina Georgieva warned that optimism about AI has energized markets and helped support the global economy. However, a sharp correction in stock prices could slow economic growth, expose vulnerabilities in the global system, and make life more difficult, particularly for developing countries.
The Bank of England said that, by some measures, stock valuations on U.S. equity markets resemble those seen near the peak of the so-called dot-com bubble, which burst in 2000.
